Back to top

Ethnographic Film

Ethnographic Film
Revised Edition

A new, thoroughly updated edition of a classic volume on visual anthropology.

November 2006
This is a print-on-demand title. Expedited shipping is not available.
180 pages | 6 x 9 | 1 figure, 19 tables |

Even before Robert Flaherty released Nanook of the North in 1922, anthropologists were producing films about the lifeways of native peoples for a public audience, as well as for research and teaching. Ethnographic Film (1976) was one of the first books to provide a comprehensive introduction to this field of visual anthropology, and it quickly became the standard reference.

In this new edition, Karl G. Heider thoroughly updates Ethnographic Film to reflect developments in the field over the three decades since its publication, focusing on the work of four seminal filmmakers—Jean Rouch, John Marshall, Robert Gardner, and Timothy Asch. He begins with an introduction to ethnographic film and a history of the medium. He then considers many attributes of ethnographic film, including the crucial need to present "whole acts," "whole bodies," "whole interactions," and "whole people" to preserve the integrity of the cultural context. Heider also discusses numerous aspects of making ethnographic films, from ethics and finances to technical considerations such as film versus video and preserving the filmed record. He concludes with a look at using ethnographic film in teaching.

  • Preface
  • Acknowledgments
  • 1. Introduction
    • Toward a Definition: The Nature of the Category "Ethnographic Film"
    • The Nature of Ethnography
    • The Differing Natures of Ethnography and Film
    • "Truth" in Film and Ethnography
  • 2. A History of Ethnographic Film
    • Background Factors
    • Prehistory: The Explorers
    • Grass
    • Scripted Fictional Films
    • Bateson and Mead in Bali and New Guinea
    • Jean Rouch
    • John Marshall
    • Robert Gardner
    • Timothy Asch
    • University of California American Indian Series
    • The Netsilik Eskimo Project
    • Australia
    • The Natives' View
    • Institutionalization of Ethnographic Film
  • 3. The Attributes of Ethnographic Film
    • The Attributes
    • Additional Principles
    • The Attributes as Dimensions
  • 4. Making Ethnographic Film
    • The Ethics of Ethnographic Filmmaking
    • An Ethnographic Film Must Be Based on Ethnographic Understanding
    • An Ethnographic Film Must Exploit the Visual Potential of Film
    • Whole Bodies, Whole Interactions, and Whole People in Whole Acts
    • Division of Labor
    • The Meaning of Real Collaboration
    • An Ethnographic Film Cannot Stand by Itself
    • Ethnographic Films from Research Footage
    • Preservation of the Film Record
  • 5. The Use of Ethnographic Films in Teaching
    • Films and Background Reading
    • Strategies
  • Appendix: A Brief Descriptive Catalog of Films
  • Bibliography
  • Index

Karl G. Heider is Carolina Distinguished Professor of Anthropology at the University of South Carolina, Columbia, where he currently serves as Associate Provost and Dean of Undergraduate Studies.


Browse the book with Google Preview »

What is "ethnographic film"? The term itself seems to embody an inherent tension or conflict between two ways of seeing and understanding, two strategies for bringing order to (or imposing order on) experience: the scientific and the aesthetic. The evolution of ethnographic filmmaking has been a continuous process attempting to reconcile this tension, to achieve a fertile synthesis. Ideally, ethnographic films unite the art and skills of the filmmaker with the trained intellect and insights of the ethnographer. This book examines the implications and the opportunities of such a collaboration; its argument rests on the fundamental assumption that film can be an important medium for the expression of the ethnographic enterprise. What is involved, though, is more demanding than a mere mechanical joining of cinematography and ethnography. There must be interpenetration of disciplines: cinematographers must accept the scientific demands of ethnography; ethnographers must adapt their expressions to the expanded visual potential of film and video. Filmmakers must think ethnographically, or scientifically; ethnographers must think cinematographically, or visually. There are great challenges on both sides, and the burgeoning vitality of ethnographic film testifies to the rewards as these challenges are met.

This book is written for people in film, for people in anthropology, and for the increasing number of people who are concerned with both, such as teachers who use these films in their classrooms. Most of it is an attempt to develop a systematic way of thinking about ethnographic film, and in particular about the "ethnographicness" of film. The issues that it treats are very close to the concerns of primatologists, sociologists, geographers, and psychologists. These neighboring specialists will have little trouble translating it into their own idiom to deal with their own interests.

We have to be very careful when talking of the differences between film and ethnography. We will find enough real problems without fostering an unnecessary rivalry. Three important points should be brought up immediately. First, the simple opposition of scientific ethnography versus artistic cinematography is misleading. Film has its share of science, both in the technology of filmmaking itself and, on another level, in the theorizing about film that has been going on for decades. On the other side, ethnography is a most impure science whose research often is carried out by a relatively private field experience. As we think about the natures of film and ethnography in the next chapters, each will shed new and interesting light on the other.

The second major point is that when the demands of ethnography and film seem to conflict and we choose to fulfill the ethnographic demands in order to have a more ethnographic film, the result will probably turn out to be a better film even by cinematographic standards. The ethnographic approach will seem strange at first in some respects to many filmmakers. But I hope to demonstrate that, in the long run, better ethnography will make for better cinematography. And third, although we are thinking about ethnographic film, and developing ways to judge ethnographicness in film, this does not imply any moral superiority for one sort of film over another. There are many worthy sorts of films. This book is about one sort, namely, ethnographic film. Likewise, there are many valid ways to look at human behavior. This book is about one way, namely, ethnography. In short, I am not presuming to make over the world film industry along the lines of ethnographic precepts; but at the same time I resist the careless expropriation of the term ethnographic to cover any film about people.

A word of explanation about terminology: I use "ethnography" and "anthropology" more or less interchangeably here. Ethnographers are anthropologists who study human behavior. They call themselves ethnographers, ethnologists, social anthropologists, cultural anthropologists, anthropologists, social scientists, or scientists, depending on the context and the audience. Some of them care very deeply about these labels and spend a lot of energy on label clarification. I don't, but I apologize for any offense caused.

A second terminological note: even though more and more of these products are shot, edited, and distributed on videotape or DVDs, I am using the old term "ethnographic film" to include film, tape, and disc. For some sorts of discussions it is necessary to distinguish among the various media. But for the purpose of this book, the single phrase "ethnographic film" will do. And in fact, these days we are often not told if a particular product originated on celluloid or tape, in analog or digital form. So, for "film," despite an occasional infelicitous phrasing, please understand "film or videotape or disc."

There are two rather different uses of film in anthropology. One is the subject of this book, that is, finished films that have been made for public showing and that present some sort of ethnographic understanding. But anthropologists also use the movie camera and the videotape recorder to make visual records of human behavior that may be analyzed in real time or may be subjected to detailed, micro-, or slow-motion analysis. These raw data are footage or film, but they are not films. After the analysis has been made, the footage may be edited into a film for public presentation of the results of the analysis, just as raw material is often incorporated into scientific papers and monographs. Gregory Bateson 's footage from Bali in the 1930s, edited into film by Margaret Mead a decade later, and Ray L. Birdwhistell's film Microcultural Incidents in Ten Zoos are two of the few examples of footage that has served both functions. The use of film (and videotape) to make visual records for behavioral analysis is a powerful tool in anthropological research. However, this still lags far behind the development of ethnographic film. But see, for example, books by Banks and Morphy (1977), Pink (2001), and Pink et al. (2004).

This book has three characteristics that are not universally admired: it is polemic, it is repetitive, and it speaks in different tongues to different audiences. It might help matters to explain why I feel that each is a useful strategy. We will never have the definitive book on ethnographic film. The whole multifaceted subject is very much in a state of flux, and we do need strong statements that organize some of the chaos and push ahead the thinking about ethnographic film. That is what I tried to do in the first edition, what many others have also tried to do, and what I continue in this new edition. I do not see this book as definitive. But it does mean to continue the debate(s) about ethnographic film.

The repetition comes from a strategy of crisscrossing exposition. The subject of ethnographic film has so many facets that it must be approached from many different directions in order to be at all exhaustive. The problem simply does not lend itself to a logical chain exposition. Thus, the same principle may be treated in one way in the theoretical chapter 1, another way in the historical chapter 2, yet another way in chapter 3, which is on attributes, and then restated in reference to ethnographic filmmakers (chapter 4) and to teachers (chapter 5). The readers' indulgence for such repetition is begged on the grounds that each reappearance will add to the development of the ideas.

Since the first edition of this book in 1976, many others have appeared on the subject (e.g., Stoller 1992, Loizos 1993, Morris 1994, MacDougall 1998, Ruby 2000, Grimshaw 2001, and El Guindi 2004, to mention just a few). We have all differed on points large and small. But in this second edition, rather than trying to concur, refute, or reconcile in detail, I have refined and updated the information and the perspectives of the first edition.

And finally, I am hoping that the various audiences of anthropologists, filmmakers, schoolteachers, and interested others can all be comfortable between the covers of a single book. No one group will want to read every page. But each audience has some concerns in common with each other audience, and in any case it is often good to know what is being said to others.

This is a fairly personal book in the sense that I draw heavily on my own experience in studying the Grand Valley Dani of Irian Jaya, Indonesia (the province has also been called Netherlands New Guinea, West New Guinea, Irian Barat, and Papua). I spent nearly three years doing research among the Dani on four trips between 1961 and 1970 (and brief visits in 1988 and 1995). I have described the Dani in an ethnographic monograph (The Dugum Dani, 1970), a case study (Grand Valley Dani, 1997), and numerous papers, as well as through two short films (Dani Sweet Potatoes and Dani Houses), and, during my first months with the Dani, I collaborated with Robert Gardner as he made the film Dead Birds. More recently I have spent another three years studying emotion and folk psychology with the Minangkabau of West Sumatra. I used video in my research but have not yet made an ethnographic film about the Minangkabau. I could never understand my own professors who had done original research but who, from some misguided sense of modesty perhaps, taught only from other people's data and refused to discuss their fieldwork. My own experience as an ethnographer as well as an ethnographic filmmaker is so directly pertinent to a book on ethnographic film that I can hardly pretend noninvolvement. Also, I know my own ethnography and my own films better than I do anyone else's. And I can criticize them with impunity.

Should a book on ethnographic film include photographs or not? When writing the first edition, I finally decided against it. It could have had photographs of the various filmmakers, cameras at hand, standing on location: interesting, but not essential. It could have "production stills" of the sort displayed outside movie theaters. These are still shots that were taken during the filming, but they rarely show scenes that are actually in the film itself. Or it could have photographs printed directly from frames in the film, but necessarily of poor technical quality by the time they reached the printed page. However, the real problem with still photographs in a book like this is that we are concerned with the problems that emerge from trying to make sense out of long sequences of moving pictures, while a still photograph is, after all, still. Ideally there would be a short film to illustrate the principles discussed here. Thirty years later I still include no photographs, but I have no doubt that in another thirty years works like this will be able to include film clips along with the text.

Inevitably a book such as this one puts great demands on the readers' imagination because of the constant reference to scenes from various ethnographic films. I have tried to keep the references limited to the films that are best known and most widely available. (The film catalog in the appendix will be of help here.) And I have tried to make the references explicit enough that even readers who cannot view the films will be able to visualize the scenes.

It has been thirty years since I wrote the first edition of this book on ethnographic film. Much has changed in the meantime. It is not just that thousands of new films have been made. More importantly, new video technology has made production much cheaper, has eliminated the need for large crews on location, and has made sync sound routine. At the same time, in ethnography itself there has been a growing concern for allowing the voices of the people to be less filtered through the outside ethnographer. With this happy conjunction of technological accessibility and theory, we have seen a proliferation of varieties of ethnographic film. Film is being used to explore the visual and aural possibilities of ethnography, taking us far beyond the traditional printed text.

Also, the last quarter century has seen the publication of a large number of books about ethnographic film, written by filmmakers, anthropologists, and others. Today many annual, biennial, and occasional conferences on ethnographic film fill our calendars. And where once the subject was hardly known and rarely practiced by anyone outside western Europe and North America, it is now approaching worldwide status.

But even after all the developments of the last decades, we are still working over some of the same old questions. The first of these is definition.

Toward a Definition: The Nature of the Category "Ethnographic Film"

What do we mean by "ethnographic film"? It is always comfortable to have a concise definition of title phrases. However, this desire for precision can often be dangerous. Of all the words that have been spilled over the definition of ethnographic film, most have treated "ethnographic" as an absolute, present-or-absent characteristic. Let us instead think of "ethnographic" as a continuously variable property of many films. Throughout this book I use the term "ethnographicness" to talk about films. The term is not particularly euphonious, but it does serve notice that "ethnographic" has a very specific meaning.

We can begin with an analogy to "tall buildings." Tallness is obviously an attribute of buildings. Just how tall a building is makes a great deal of difference to an architect, a builder, and an occupant. They all need to have ways to measure and discuss tallness. But it would make no sense at all to attempt to define tall buildings so that some buildings were "tall" and some "not tall." For decades we have spent much energy trying to define ethnographic film in this way. But there is a different approach that is more useful. As long as we phrased the questions in the form "What is an ethnographic film?" or "Is X an ethnographic film?" we were assuming the existence of a bounded category. We had to direct our energies to discovering the boundaries, and we had as our goal the definition of a set of boundary criteria that would allow us to mark off some films as "ethnographic" and, at least implicitly, the rest as "not ethnographic." Certainly the underlying problems are real, but the terms of the inquiry have been sterile. I would suggest a moratorium on such questions.

And now even the tall-buildings analogy becomes inadequate, for "tall" is a single or unidimensional measure. As we talk about films, we must deal with many aspects.

Some have tried to dodge the issue by resorting to broader terms like "documentary film" or "nonfiction film"; some develop specialized terms like "observational cinema" or "transcultural cinema." All of these labels have their uses, but the inherent problem of trying to defend boundaries remains.

So a useful approach is to look for the various attributes, or dimensions, that effect ethnographicness in films. Adopting this strategy, we should ask, "What film features may be more or less ethnographic?" and "In what ways is this film ethnographic?" The idea, then, is not to define an ethnographic box-category but to make explicit those features that contribute to the ethnographicness of films.

There are two overriding considerations:

  1. How closely can films approach the highest standards and goals of ethnography?
  2. How can films present information that written ethnographies cannot?

To resolve the apparent paradox of these two considerations—being ethnographic and transcending ethnography—we can rephrase them as follows:

How can the (visual capability of) film complement the (lexical capability of) ethnography?

The truism that "a single picture is worth ten thousand words" can be inverted as "a single word can be worth ten thousand pictures." Depending on the situation, either may be true. The challenge of ethnographic film is to develop ways of thinking about film that will make it more ethnographic.

Even though we refuse to define ethnographic film, we must make explicit those attributes that make some films more ethnographic than others. This will make possible a criticism of films from an ethnographic standpoint. More important for the future, it will allow us to explain clearly to ethnographers and to filmmakers how to make films that are more ethnographic and thereby more valuable to anthropology.

In evaluating the degree of ethnographicness in any film, or in designing film projects that are maximally ethnographic, we need to consider a number of attributes, some of which emerge from primarily ethnographic constraints, and some of which emerge from cinematographic constraints. We must stress early and often that when we are talking about "ethnographic film," ethnography must take precedence over cinematography. If ethnographic demands conflict with cinematographic demands, ethnography must prevail.

There are lots of ways to skirt this issue, but it really must be faced head-on. In ethnographic film, film is the tool and ethnography the goal. I have talked with filmmakers who see film as somehow an end in itself, and they reject this idea as demeaning to the art of cinematography. I think that they misunderstand film. It is always a tool for something, and in talking about ethnographic film we are just making that something more explicit than usual. But certainly there is an opposite danger: that anthropologists will see film as "only a tool" and not give it the respect and attention that a fine tool demands. Film-the-tool does some things magnificently well, but for other things it is quite inadequate. We have to understand enough so that we do not try to saw wood with a hammer.

Sometimes this matter is phrased as an inevitable contradiction between art and science, with filmmakers arguing the case for art, and anthropologists the case for science. But this is a distracting approach. The analogy with ethnography-as-literature is pertinent here. We do not demand that an ethnography be written in great literary style. However, when poor writing obscures the ethnographic point, we rightly object. Similarly, although we need not hold ethnographic films to the highest cinematographic standards, a minimal cinematographic competence is required in order that the film communicate at all. In fact, however, it seems safe to say that films that are cinematographically incompetent are also ethnographically incompetent (even when made by an ethnographer). Most of the films we use are far more successful in a technical cinematographic sense than they are in any ethnographic terms. The main problem, and the one to which this book is devoted, is to show filmmakers as well as ethnographers what it means to say that films are ethnographic.

Ethnographic film must be judged in relation to ethnography, which is, after all, a scientific enterprise. In some sense one could argue that all films are "ethnographic": they are about people. Even films that show only clouds or lizards have been made by people and therefore say something about the culture of the individuals who made them and who use them. Many films that have little pretension to ethnographicness are nevertheless of great interest to the ethnographer. I personally feel that The Last Picture Show, about the high school class of 1952 in a small Texas town, is a statement that captures the culture of my own high school class of 1952 in Lawrence, Kansas. Likewise, The Harder They Come (about Jamaica), Scenes from a Marriage (about middle-class Swedish marriage), and Tokyo Story (about an extended Japanese family) all present important truths about cultural situations. And I have written elsewhere of how Indonesian feature films are shaped by Indonesian cultural norms (1991a, 1991b). Also, in my cultural anthropology textbook (2004), I have suggested using fiction films that deal with topics of cultural anthropology. As statements (native statements, in fact) about culture, these films are important, and they could serve as raw data or documents in certain kinds of ethnographic research. I am tempted to call them more than just "raw data" and to think of them as "naive ethnography." They have ethnographic import without attempting the science of ethnography. They are good entertainment, but they are also certainly worthy of serious consideration. I have sometimes used such films in my anthropology classes. However, here I want to consider films that are more specifically and intentionally ethnographic.

The Nature of Ethnography

The most important attribute of ethnographic film is the degree to which it is informed by ethnographic understanding. It would be difficult to define "ethnography" in a few words, but we must discuss those features of ethnography that are most relevant to ethnographic filmmaking (and here I shall talk about a general, prototypical sort of ethnography).

First, ethnography is a way of making a detailed description and analysis of human behavior based on a long-term study on the spot. A "come-in-shooting-and-get-out-fast" approach and an intuitive-aesthetic appreciation of behavior and people are other sorts of approaches, and although they may well result in beautiful films, these must inevitably be ethnographically shallow. The Nuer is a good example of the limitations of the last approach. It was shot in the late 1960s among a group of Nilotic cattle herders who were living in Ethiopia. Although the filmmakers did spend several months on location and did capture the pace of life in a cattle camp, they missed significant aspects of Nuer ethnography. There is a sequence of the boys' initiation that is dramatic and moving, but from any ethnographic standpoint it is incomplete. On the most obvious level, the two boys are shown leaving boyhood, but their entrance to manhood is omitted except for a slight reference in the narration. The specific important steps in Nuer initiation ceremonies and the structural relation of these ceremonies to rites of passage in general, both of which are well known in ethnographic literature, are hardly touched on in the film. The initiation sequence could have been shaped by these understandings without sacrificing any of the aesthetic qualities of the film. The result would have been stronger ethnographically.

Another essential feature of ethnography is that it relates specific observed behavior to cultural norms. Many documentary films devote much time to the portrayal of an individual person or event but fall short of the cultural step, putting those data into a cultural context. General cultural statements are especially challenging since they must almost inevitably be made in words, whereas film is by nature specific and visual. The difficulty of saying no or asking a question through purely visual means in film or photograph has often been remarked. A photographic generalization is almost as hard to achieve. One of the rare attempts occurs in The Nuer, where several quick sequences illustrate various tobacco pipes, ornaments, and scar patterns. The most common solution, used in most films, is to have the narrator read words that put the specific visual images into generalized cultural context. But this forces the film to take on some of the quality of a book, at the expense of the purity of its imagery. There is a real temptation to load too much information into the narration, further weakening the "filmicness" of the film, and at times even contradicting the visual information.

A third basic principle of ethnography is holism. To some degree, things and events must be understood in their social and cultural context. From this principle come the related dicta of "whole bodies," "whole people," "whole interactions," and "whole acts." They emphasize the ethnographic need to present bodies, personalities, and behavior in context. This does not mean that isolated close-ups of body parts (especially faces), fleeting glimpses of impersonalized people, and fragmented representations of behavioral events can never be used, but it recognizes the importance of more holistic shooting.

Like holism itself, "whole people in whole acts" is not a dictum to be followed slavishly. It does not mean that we describe or film everything about everything. A twelve-hour camera's-eye view of life in the village square would not be ethnographically effective. But the holistic principle must be kept in mind as a corrective principle for making films more ethnographic.

Another major feature of ethnography is the goal of truth. On some philosophical level it may be argued that this is a hopelessly naive goal and that reality cannot be truthfully represented. But for our present purposes we can usefully hold that accuracy and truth are essential to ethnography, that some accepted conventional distortions of reality occur in the translation of the living act onto the printed page, and that ethnographers are fairly well aware of the conventions of distortion and are fairly well agreed on what constitutes illegitimate (by name, "dishonest" or "unscientific") distortion.

The conventions of cinematographic honesty are quite different. Cinema has developed primarily as a medium for imaginative statements in which questions of scientific-type accuracy are often irrelevant. Much of what is taught in film schools is how to translate or distort reality for aesthetic effect. These techniques include selective composition of shots, staging acted scenes, editing for continuity effect, and using sound recorded in other contexts. Some of these reality-distorting techniques are inevitable in even the most scrupulously ethnographic films. But in order to understand the ethnographicness of a film, we need to know how much and to what degree reality was distorted. And in making ethnographic films, we can ask that distortions are kept to a minimum and used for ethnographic purposes, not for merely cinematographic reasons. For example, when editing Dani Houses and Dani Sweet Potatoes, I could have used vaguely appropriate sound recorded two years before I shot my own footage. But I chose not to use this wild sound or any sound other than the narration. As a result, the two films seem tedious and empty to some viewers who expect to be continuously entertained by the sound track. But I had decided, on ethnographic grounds, that the inappropriateness of the wild sound would override the viewing pleasure of audiences accustomed to more technically elaborate films.

Choices of this sort must be made deliberately, not by default. For example, when Donald and Ronald Rundstrom and Clinton Bergum were making The Path, they wanted to describe the flow of movement and the use of kinesic energy in the Japanese tea ceremony. To show this with the greatest effectiveness, they shot and edited in a way that lost the casual social gossipy nature of the tea ceremony. The film represents only one side of the reality of the ceremony. But the decisions of what to depict and what to omit were deliberate ethnographic decisions.

These, then, are the ethnographic demands for a general-use ethnographic film, that is, one that is relatively self-explanatory for casual use but demands an accompanying written ethnography for more serious use and deeper understanding. It is difficult to imagine a film that could carry enough ethnographic contextualization and generalization to be fully self-sufficient. At the other extreme, we are now beginning to see some videos that use only synchronous sound, without any generalizing narration. Ethnographically, such films can be extremely effective when used in close connection with a written ethnography or when presented by a well-prepared instructor/informant. But even though they may entirely relegate the generalizing to the written word, the degree to which these films are judged ethnographic must still depend on the degree to which they satisfy the other requirements of ethnographic film discussed above.

To define "ethnographic film" requires either one sentence or an entire book. The sentence is "Ethnographic film is film that reflects ethnographic understanding." This book is an exploration of the nature of ethnographic film. Whatever it is, it is more than the simple sum of ethnography plus film.

It is easy to show the faults in those ethnographic films that are even less than the sum of the two: films made by an ethnographer who happens to take a movie camera into the field, or films made by a filmmaker who happens to wander among an exotic tribe. Often even such films have some documentary value, but they inevitably are monumental missed opportunities. The answer is not simply for a filmmaker to take a course or a degree in anthropology; nor is it enough for an anthropologist to enroll in film school. This is what I mean by emphasizing that ethnographic film is more than the mechanical joining of the two. The effort of thinking cinematically about ethnography or thinking ethnographically through film results in a new and different understanding of each of these disciplines.

Both film and ethnography involve particular ways of viewing the world. Ethnographic film should represent the best of both ethnography and film. This can be done by respecting the constraints of each discipline and taking advantage of the opportunities and insights of each. All this is just saying—in abstract, programmatic phrases—that which the rest of this book makes explicit.

The Differing Natures of Ethnography and Film

There are profound differences between ethnography and film that are corollaries of the obvious differences between word-on-paper versus photo-on-celluloid or image-on-videotape. Having said that the differences begin with word versus picture, I will examine the consequences of these differences, and in particular how all this affects the way in which people go about the two enterprises.

The ethnographic research that culminates in a book and the filmmaking that culminates in a film are quite different enterprises that can be schematized as follows:

The ethnographer The filmmaker
Begins with theoretical problems and research plans Begins with an idea and a script
Gathers data by making observations and asking questions Shoots footage
Analyzes data Edits footage
Writes and rewrites  
Produces a written report Produces a film

Cinematography makes irreversible choices at the very beginning. The finished film can contain only those images that were shot at the beginning. Especially in ethnographic filmmaking, where the shooting is done in the field and is completed before the editing begins, the editing stage is one of manipulating a finite amount of set material. Scenes can be shortened, thrown out, or their sequence juggled, and titles and narration can be added. But, practically speaking, new footage cannot be created at the editing table.

Imagine for a moment a comparable situation in ethnography. The ethnographer spends a few months in the field, writing down observations in notebooks. Then she returns home, makes a photocopy of the notebook, and with scissors and paste proceeds to fashion an ethnography out of the sentences and paragraphs she wrote down in the field. She can take sentences from different pages of the notebook and juxtapose them into new paragraphs. But she cannot write any new words or sentences.

In reality the ethnographer writes and rewrites, analyzes and reanalyzes; in short, she composes and recomposes words into sentences, and sentences into paragraphs. The final ethnography hardly contains a single phrase as originally set down in the first field record. On the other hand, except for a few printed titles, the final ethnographic film contains only images that were originally photographed. The basic difference in the way in which understanding enters the process is dramatically illustrated by the fact that when the footage has been shot, someone other than the photographer can (and usually does) edit it into the finished film, but it would be almost impossible to write an ethnography from someone else's field notes.

This difference between editing (the film) and rewriting (the book) has several implications. First, whereas the thorough understanding of what is going on often emerges only at the end of the ethnographic process, it must precede the filmmaking. Of course, the ethnographer must begin gathering data with some research plan in mind. But actually, many of the most important data are gathered on the fringe, so to speak, by means other than the preestablished research procedures. The peripheral vision of the anthropologist is often an extremely important research tool. We rarely talk about it or systematize it, but we do use it. Also, as anthropologists come to understand the behavior of a people better, they can retrieve data from their own memories or from the memories of their informants. Such convenient hindsight is of no help to the cinematographer. Time past cannot be refilmed. A ceremony is filmed on the basis of whatever understanding (and luck) the filmmakers had at the moment of filming. (And I would insist that "lucky shots" are usually the rewards of understanding.) Later, when the filmmakers know more, they can film a different ceremony, and they can cut and shuffle the old footage. They can edit, but they cannot rewrite.

Ethnographic understanding emerges from the analysis, and an ethnography is only as good as the analysis. But an ethnographic film can only be as good as the understanding that precedes the filmmaking. Or, put another way, the degree to which a film is ethnographic depends on the degree to which prior ethnographic understanding has informed the filmmaking. This is the most basic message of this entire book, and I shall repeat it often in different ways.

Filmmaking is not quite so rigidly locked into first impressions as the preceding lines might suggest, but the qualifications only strengthen the argument. Often in ethnographic filmmaking, we find a wordy narration used as a bridge, or a bandage, to patch over a gap in the footage where some sequence has not been fully filmed. Often this gap is the result of ignorance, where the cameraman has not understood well enough what was going on to be able to anticipate and to shoot satisfactory footage. The inability to anticipate meant that the camera batteries went dead, or the camera was unloaded at the crucial moment (I shot an important scene for Dead Birds with an unloaded camera), or the cameraman was not even present. But it is significant, in the context of this argument, that the filmic gap is frequently closed by words. When this is done, then the filmicness of the film is diminished, and the product edges toward a spoken book. We have already insisted that ethnographic films be maximally ethnographic; but ethnographic films should also be maximally visual.

"Truth" in Film and Ethnography

A basic problem, already mentioned in this chapter, that runs through all considerations of ethnographic film concerns the nature of truth. Filmmakers and ethnographers, when they think of it at all, take quite different positions on truth. Certainly everyone subscribes to truth. No one really advocates untruth. (Those who deny the very possibility of truth remove themselves from this discussion.) But filmmakers can comfortably take the artists' position that they manipulate reality through a series of falsehoods in order to create a higher truth. This appeal to higher truth has been made in more or less epigrammatic form, by all sorts of artists. Anthropologists, as scientists, assume that they must challenge the legitimacy of these facilitating lies. In science, the end cannot justify the means: results are only as sound as the methodology that produces them. Nice words. But if we look a bit more closely, we see that anthropologists have their own methodological conventions for reaching truth. And we should be rather explicit about them.

If one were to ask anthropologists to name the five best ethnographies, or the five that they liked best, one would have no trouble getting lists from most anthropologists. Actually, "five best" might be too much of an American-specific concept, and one might have to rephrase the question for anthropologists of other cultures—for example, "five really first-rate ethnographies." But lists would still be forthcoming. If one next asked the same anthropologists if truth was important to ethnography, the response would be a nearly unanimous yes. But then if one asked for a list of the five truest, or most accurate ethnographies, there would be trouble. People would hesitate and ask for clarification, and if they produced a list at all, it would not overlap much with their list of the "five best."

Assuming that I am correct about the thinking of my anthropological colleagues, what is going on here? One important difference between the five best and the five truest ethnographies would probably be that the "truest" were safe catalogs of trait lists, while the best were works that did not attempt any sort of exhaustive coverage. Rather, they would take a selection of data and interpret these data in a particularly enlightening and convincing manner. The truth perhaps, but certainly not the whole truth.

In short, we do not really expect any ethnography (or film) to say everything about a subject. This means that there must be selection and there must be omission. Therefore, the value of an ethnography or a film cannot be judged on the basis of whether or not it has omitted things. Rather, it must be judged on the appropriateness of what has been included and how the content has been handled.

All this must seem obvious, but in actual practice it is easy to forget. For example, many anthropologists criticized Robert Gardner's film Dead Birds on the grounds that it showed only the men's side of Dani life. When it was pointed out that many women's activities were in fact shown, the criticism was modified to "Well, he didn't show enough of the women." Gardner's later film Rivers of Sand concerns the role of women among the Hamar of Ethiopia and is dominated by the statements of one Hamar woman. Now the criticism is reversed: Gardner doesn't present the men's side. It would be small wonder if Gardner felt frustrated by these criticisms. Underlying them is the feeling that somehow a film should cover everything. But that is a feeling held by anthropologists who would not make such a demand of ethnographies.

There is another convention that anthropologists accept in written ethnographies but often challenge in films. That convention is the common device of reconstructing an account of an event from diverse data in order to make a single, reasonable, typical, "true" account. When audiences learn that the giraffe hunt in The Hunters was constructed in the editing room out of scenes from several different hunts, or that the battle sequence in Dead Birds combined footage shot at different battles, they often feel betrayed and lose confidence in the total description of a Bushman hunt or a Dani battle.

But, in fact, a comparable construction is done in ethnography—for example, in Bronislaw Malinowski's Argonauts of the Western Pacific (1922). This is one of the earliest and still one of the most respected ethnographies. It would certainly appear in nearly every anthropologist's "five best ethnographies" list, discussed earlier.

The middle section of that book describes a long trading expedition undertaken by the Trobriand Islanders from their homes in Sinaketa, in the north, to Dobu, an island to the south. Now, Malinowski had not actually participated in any such expedition. His account is reconstructed from brief observations of somewhat similar events and from accounts that he gathered from his Trobriand informants. His defense of this technique is worth quoting at some length, because in so many ways Malinowski laid the foundation for ethnographic fieldwork and is still a powerful role model for ethnographers:

In the twelve preceding chapters, we have followed an expedition from Sinaketa to Dobu.... As I have seen, indeed followed, a big uvalaku expedition from the South to the Trobriands, I shall be able to give some of the scenes from direct impression, and not from reconstruction. Such a reconstruction for one who has seen much of the natives' tribal life and has a good grip over intelligent informants is neither very difficult nor need it be fanciful at all. Indeed, towards the end of my second visit, I had several times opportunities to check such a reconstruction by witnessing the actual occurrence, for after my first year's stay in the Trobriands I had written out already some of my material. As a rule, even in minute details, my reconstructions hardly differed from reality, as the tests have shown. None the less, it is possible for an ethnographer to enter into concrete details with more conviction when he describes things actually seen. (Malinowski 1922, 376)

This sort of reconstruction of an event in written description is an accepted convention in ethnography. It is not exclusively used, of course. Malinowski himself brought in rich specific case studies, or anecdotes, as well as these generalized reconstructions. The point is that the general reconstruction, when used properly, is a legitimate ethnographic descriptive device. But there are differences between print and film uses of this convention. A good example is the housebuilding of the Dani. I have described Dani construction in print (1970, 261-263) and in film (Dani Houses, 1974). During the years that I lived in West New Guinea with the Dani, I observed (and often participated in) housebuilding, but I never saw all the steps in building any one house. I formed a complete picture of Dani construction from countless isolated observations of Dani at work. I am now somewhat embarrassed to realize that I never actually wrote these qualifications in my ethnography. And my account is phrased in very general terms. It begins: "The main part of the construction is done by the men and boys who will live in the men's house. Chopping of lumber and actual construction is done only by men; women carry wood and thatch grass to the site..." (1970, 261).

I did not tell the reader where I got my information. But presumably most readers would understand this as a general account. And, in contrast, when I later discussed "magic associated with housebuilding," I was very careful not to claim these data as general, specifying the names of the compound sites: "At Anisimo a knot of grass... At Biem, dried banana leaf was wrapped around the base... At Musanima..." (1970, 263). In the film Dani Houses, I used footage of that instance of banana leaf magic mentioned in this quotation. However, although in the book it is clear that I observed that banana leaf magic only once, I did not put this information into the narration of the film. And, in fact, most viewers would probably assume that such banana leaf magic was a standard part of Dani housebuilding. On the surface this is a minor ethnographic detail, but such situations are multiplied many times in the course of any ethnographic film, and the problem of truth becomes a major consideration.

Also, while shooting the film Dani Houses, I had missed several steps in the sequences. But during the editing, I made the decision not to edit diverse shots together for continuity but rather to keep them separate, in the order in which they had occurred. For example, the first sequence in the film follows the construction of two pigsties. I had footage of women carrying the thatch grass to the construction site, but I did not get footage of the grass being plucked. For a later sequence, there is footage of both the plucking and the carrying to the site of a round house. I decided to use the plucking footage in the second sequence, where it belonged. But that means that we first see thatch grass appear without knowing how it was gathered. It would have made for a smoother, more logical film to have used the plucking footage in the first sequence. But since that footage was of different people in a different place, I held out for accuracy.

There was yet another dilemma in Dani Houses. I had filmed half the construction of a men's round house and then, through some circumstances not relevant to the housebuilding itself, I missed the completion of that house. Later I filmed the second half of the construction of a women's round house. It would have been possible, by choosing proper shooting angles, and through editing, to present the two houses as one. Instead I deliberately emphasized that they were separate and mentioned the differences between men's and women's houses.

These sorts of specific, demystifying approaches are becoming more common in ethnography as well as in ethnographic film. We have seen that Malinowski, writing in the 1920s and 1930s, used both the specific illustration and the generalized description. My own 1970 monograph is somewhat more generalizing than Malinowski's works, while my films (finished several years after the monograph) represent more uncompromising specifics. Of course, words lend themselves to generalizing statements more easily than does film, whereas the specificity of the film image makes filmic generalizations less satisfactory and more manipulative.

In written accounts we usually have some idea of the extent to which a description is specific, reconstructed, or generalized. The choice of verb form and the use of parenthetical expressions, footnotes, and explanatory paragraphs offer opportunities to qualify or support written data. We usually do not get comparable information in films. It is more difficult to slip parenthetical information into a film. When a description must be seen from beginning to end at a predetermined speed, without pause or backtracking, there is no place for a footnote. Of course, it would still be possible to include qualifying information in a film, but it would be more intrusive and the conventions of filmmaking are against it.


Available for Kindle
Available on Google Play
Available on Kobo
Available for Nook
Available on the Apple Store

This book may also be available on the following library platforms; check with your local library:
3M Cloud Library/bibliotheca
UPCC/Project Muse